The long-term goals of the proposed research are to improve the acquisition, reconstruction, and extraction of quantitative information from emission tomographic images of challenging radionuclides and to assess these improvements using task-dependent criteria. During the current project period, we have concentrated our efforts on complex radionuclides, e.g., 111In and 67Cu, as well as on simultaneous dual- radionuclide imaging of 111In and 99mTc. During the next project period, we will refine our research with these radionuclides;however, we will also extend the work in three new directions in order to address important technical challenges in pre-clinical emission tomographic imaging of mice, bremsstrahlung imaging of beta emitters used for radionuclide therapy, and reduction of systematic errors due to physical factors that currently limit the accuracy and precision of treatment planning in targeted radionuclide therapy. The imaging tasks that we will consider involve estimation of activity concentration in tumors or in regions of infection. We will measure, using simulated data as well as phantom, animal, and/or patient data, the improvements in task performance resulting from our new methodologies, and compare the performance achieved to theoretical bounds. We will consider clinical tasks related to treatment planning for radiotherapy of tumors expressing somatostatin receptors, as well as to treatment of hepatic tumors and metastases with 90Y- labeled microspheres, and to the diagnosis of osteomyelitis. We will extend our SPECT collimator design work to the case of simultaneous dual-radionuclide imaging, and to address the unresolved issue of whether collimator optimization on the basis of projection datasets is sufficient, or whether the collimator should be jointly optimized with the reconstruction algorithm. We will also develop and assess a reconstruction technique for bremsstrahlung SPECT, as well as a model- based procedure for correcting SPECT-CT and PET-CT images for the "partial volume effect." Finally, we will compare the performance of a Monte-Carlo based iterative reconstruction algorithm to that of a quantitative planar imaging approach for the diagnosis of osteomyelitis using dual-tracer data. PUBLIC HEALTH RELEVANCE: The proposed research will address technical challenges which complicate emission tomographic imaging of several radionuclides which are of value for diagnosis, treatment planning, and therapy of cancer, as well as for diagnosis of bone infection. The insights and improved imaging techniques which will emerge from this research will lead to better diagnosis and management of these diseases.